We should search git programs only in the highest-priority location.
The old code added the directories "argv_exec_path",
"getenv(EXEC_PATH_ENVIRONMENT)", and "system_path(GIT_EXEC_PATH)" to
PATH. The same order is implemented in git_exec_path(), which returns
the highest priority location to search for executables. If the user
explicitly overrides the default location (by --exec-path or
GIT_EXEC_PATH) we can expect that all the required programs are there.
It does not make sense that only some of the required programs are
located at the highest priority location and other programs are picked
up from a lower priority exec-path. If exec-path is overridden
a complete set of commands should be provided, otherwise several
different versions might easily get mixed, which is likely to spread
confusion.
Accessing the location with highest priority only is also required for
testing of executables built with RUNTIME_PREFIX. Calling
system_path(GIT_EXEC_PATH) is avoided if a higher-priority location is
provided, which is the case for the tests. The call to system_path()
must be avoided, if RUNTIME_PREFIX is set, because the call would fail
if the executable is not installed at its final destination. But we
test before installing.
Programs that use git_config need to find the global
configuration. When runtime prefix computation is enabled, this
requires that git_extract_argv0_path() is called early in the
program's main().
This commit adds the necessary calls in the programs that use
git_config.
If the user specified the full path to git-upload-pack as the -u option to
"git clone" when cloning a remote repository, and git was not on the default
PATH on the remote machine, git-upload-pack was failing to exec
git-pack-objects.
By making the argv[0] path (if any) available to setup_path(), this will
allow finding the "git" executable in the same directory as
"git-upload-pack". The default built in to exec_cmd.c is to look for "git"
in the ".../libexec/git-core" directory, but it is not installed there (any
longer).
Much the same applies to invoking git-receive-pack from a non-PATH location
using the "--exec" argument to "git push".
[ spr: split Steve's original commit into two commits. ]
Signed-off-by: Steve Haslam <shaslam@lastminute.com>
Signed-off-by: Steffen Prohaska <prohaska@zib.de>
This commit moves the code that computes the dirname of argv[0]
from git.c's main() to git_set_argv0_path() and renames the function
to git_extract_argv0_path(). This makes the code in git.c's main
less cluttered, and we can use the direname computation from other
main() functions too.
[spr: split Steve's original commit and wrote new commit message. ]
Signed-off-by: Steve Haslam <shaslam@lastminute.com>
Signed-off-by: Steffen Prohaska <prohaska@zib.de>
This commit modifies system_path() to compute the prefix at runtime if
configured to do so. If RUNTIME_PREFIX is defined, system_path() tries
to strip known directories that executables can be located in from the
path of the executable. If the path is successfully stripped it is used
as the prefix. For example, if the executable is "/msysgit/bin/git" and
BINDIR is "/bin", then the prefix is computed as "/msysgit".
We report an error if the runtime prefix computation fails, which can
happen if the executable is not installed at a known location. The user
should know that the global configuration is not picked up, because this
can cause unexpected behavior. If we explicitly want to ignore system
wide paths, we can set the environment variable GIT_CONFIG_NOSYSTEM, as
our tests do.
The implementation requires that argv0_path is set up properly, which is
currently the case only on Windows. argv0_path must point to the
absolute path of the directory of the executable, which is verified by
two calls to assert(). On Windows, the wrapper for main() (see
compat/mingw.h) guarantees that this is the case. On Unix, further work
is required before RUNTIME_PREFIX can be enabled.
MSYS' automatic path conversion causes problems when passing paths as
defines ('-D' arguments to the compiler). MSYS tries to be smart and
converts absolute paths to native Windows paths, e.g. if MSYS sees
"/bin" it converts it to "c:/msysgit/bin". But we want completely
unmodified paths; e.g. if we set bindir in the Makefile to "/bin", the
define BINDIR shall expand to "/bin". Conversion to absolute Windows
path will takes place later, during runtime.
This commit adds a workaround by replacing "/" with its octal
representation "\057", effectively hiding the path from MSYS' path
conversion mechanism. MSYS does no longer see the absolute path and
therefore leaves it alone.
This concept was retired by 77882f6 (Retire diffcore-pathspec.,
2006-04-10), more than 2 years ago.
Signed-off-by: Yann Dirson <ydirson@altern.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
5723fe7 (Avoid cross-directory renames and linking on object creation,
2008-06-14) changed the call to use link() directly instead of through a
custom wrapper, but forgot that it returns 0 or -1, not 0 or errno.
Signed-off-by: Thomas Rast <trast@student.ethz.ch>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
There is currently no call to git_config at the start of cmd_archive.
When creating tar archives the core config is read as a side-effect of
reading the tar specific config, but this doesn't happen for zip
archives.
The consequence is that in a configuration with core.autocrlf set,
although files in a tar archive are created with crlf line endings,
files in a zip archive retain unix line endings.
Signed-off-by: Charles Bailey <charles@hashpling.org>
Acked-by: René Scharfe <rene.scharfe@lsrfire.ath.cx>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In the unusual case when there is no commit message, gitweb would
output an uninitialized value warning.
Signed-off-by: Joey Hess <joey@kitenet.net>
Acked-by: Jakub Narebski <jnareb@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The only part of the hunk header that we can change is the "TEXT"
portion. Additionally, a few grammatical errors have been corrected.
Signed-off-by: Garry Dolley <gdolley@ucla.edu>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Apparently do_switch() tolerates the lack of escaping in less
funky branch names. For the really strange and scary ones, we
need to escape them properly. It strangely maintains compatible
with the existing handling of branch names with spaces and
exclamation marks.
Reported-by: m.skoric@web.de ($gmane/94677)
Signed-off-by: Eric Wong <normalperson@yhbt.net>
Not doing so caused the "Malformed network data" error when a directoy
was deleted and replaced with a copy from an older version.
Signed-off-by: Alec Berryman <alec@thened.net>
Acked-by: Eric Wong <normalperson@yhbt.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Textual diff output for unmerged paths was too eager to give condensed
combined diff. Even though "diff -c" (and "diff-files -c -p") is a
request to view combined diff without condensing (otherwise the user would
have explicitly asked for --cc, not -c), we showed "--cc" output anyway.
0fe7c1d (built-in diff: assorted updates, 2006-04-29) claimed to be
careful about doing this, but its breakage was hidden because back then
"git diff" was still a shell script that did not use the codepath it
introduced fully.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If we came from git.c the first arg would be "archive".
"git-archive" isn't a bug because cmd_archive() doesn't check
the first arg.
Signed-off-by: Heikki Orsila <heikki.orsila@iki.fi>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
User notifications are presented as 'git cmd', and code comments
are presented as '"cmd"' or 'git's cmd', rather than 'git-cmd'.
Signed-off-by: Heikki Orsila <heikki.orsila@iki.fi>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The object oriented version of File::Temp is a rather new incarnation it
seems. The File::Temp man page for v5.8.0 says "(NOT YET IMPLEMENTED)" in
the 'Objects' section. Instead of creating a file with a unique name in
the system TMPDIR, we can create our own temporary file with a static
name and use that instead.
Signed-off-by: Brandon Casey <casey@nrlssc.navy.mil>
Tested-by: Tom G. Christensen <tgc@statsbiblioteket.dk> on RHEL 3, Perl 5.8.0
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Some versions of perl complain when 'STDERR' is used as the third argument
in the 3-argument form of open(). Convert to the 2-argument form which is
described for duping STDERR in my second edition camel book.
Signed-off-by: Brandon Casey <casey@nrlssc.navy.mil>
Tested-by: Tom G. Christensen <tgc@statsbiblioteket.dk> on RHEL 3, Perl 5.8.0
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit 3632cfc24 makes the same change for Darwin; however, the problem
also exists on FreeBSD.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This augments 3632cfc24 (Use compatibility regex library on Darwin,
2008-09-07), which already carries a "Tested-by" statement for AIX,
but that test was actually done with this patch included.
Signed-off-by: Johannes Sixt <johannes.sixt@telecom.at>
Tested-by: Mike Ralphson <mike@abacus.co.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The standard libc regex library on OSX does not support alternation
in POSIX Basic Regular Expression mode. This breaks the diff.funcname
functionality on OSX.
To fix this, we use the GNU regex library which is already present in
the compat/ diretory for the MinGW port. However, simply adding compat/
to the COMPAT_CFLAGS variable causes a conflict between the system
fnmatch.h and the one present in compat/. To remedy this, move the
regex and fnmatch functionality to their own subdirectories in compat/
so they can be included seperately.
Signed-off-by: Arjen Laarhoven <arjen@yaph.org>
Tested-by: Mike Ralphson <mike@abacus.co.uk> (AIX)
Tested-by: Johannes Sixt <johannes.sixt@telecom.at> (MinGW)
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Perl 5.8.0 ships with File::Temp 0.13, which does not have the new()
interface introduced in 0.14, as pointed out by Tom G. Christensen.
This modifies Git.pm to use the more established tempfile() interface
and updates 'git svn' to match.
Signed-off-by: Marcus Griep <marcus@griep.us>
Acked-by: Eric Wong <normalperson@yhbt.net>
Tested-by: Tom G. Christensen <tgc@statsbiblioteket.dk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This should make the output more readable (by default using diff -u)
when some tests fail.
Also changed the diff order from "current expected" to "expected
current".
Signed-off-by: Miklos Vajna <vmiklos@frugalware.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* np/maint-safer-pack:
fixup_pack_header_footer(): use nicely aligned buffer sizes
index-pack: use fixup_pack_header_footer()'s validation mode
pack-objects: use fixup_pack_header_footer()'s validation mode
improve reliability of fixup_pack_header_footer()
pack-objects: improve returned information from write_one()
The logic to checkout a different commit implements the safety to never
lose user's local changes. For example, switching from a commit to
another commit, when you have changed a path that is different between
them, need to merge your changes to the version from the switched-to
commit, which you may not necessarily be able to resolve easily. By
default, "git checkout" refused to switch branches, to give you a chance
to stash your local changes (or use "-m" to merge, accepting the risks of
getting conflicts).
This safety, however, had one deliberate hole since early June 2005. When
your local change was to remove a path (and optionally to stage that
removal), the command checked out the path from the switched-to commit
nevertheless.
This was to allow an initial checkout to happen smoothly (e.g. an initial
checkout is done by starting with an empty index and switching from the
commit at the HEAD to the same commit). We can tighten the rule slightly
to allow this special case to pass, without losing sight of removal
explicitly done by the user, by noticing if the index is truly empty when
the operation begins.
For historical background, see:
http://thread.gmane.org/gmane.comp.version-control.git/4641/focus=4646
This case is marked as *0* in the message, which both Linus and I said "it
feels somewhat wrong but otherwise we cannot start from an empty index".
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Two patches 9907721 (templates/Makefile: don't depend on local umask
setting, 2008-02-28) and 96cda0b (templates/Makefile: install is
unnecessary, just use mkdir -p, 2008-08-21) tried to prevent an overtight
umask the builder/installer might have from screwing over the installation
procedure, but we forgot there was another source of trouble. If the
person who checked out the source tree had an overtight umask, it will
leak out to the built products, which is propagated to the installation
destination.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
"git rebase --continue" issued after git rebase being stop by "edit"
command is trying to amend the last commit using stage changes. However,
if the last commit is not the commit that was marked as "edit" then it
can produce unexpected results.
For instance, after being stop by "edit", I have made some changes to
commit message using "git commit --amend". After that I realized that
I forgot to add some changes to some file. So, I said "git add file"
and the "git rebase --continue". Unfortunately, it caused that the new
commit message was lost.
Another problem is that after being stopped at "edit", the user adds new
commits. In this case, automatic amend behavior of git rebase triggered
by some stage changes causes that not only that the log message of the
last commit is lost but that it will contain also wrong Author and Date
information.
Therefore, this patch restrict automatic amend only to the situation
where HEAD is the commit at which git rebase stop by "edit" command.
Signed-off-by: Dmitry Potapov <dpotapov@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
User notifications are presented as 'git cmd', and code comments
are presented as '"cmd"' or 'git's cmd', rather than 'git-cmd'.
Signed-off-by: Heikki Orsila <heikki.orsila@iki.fi>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If git rebase interactive is stopped by "edit" command and then the user
said "git rebase --continue" while having some stage changes, git rebase
interactive is trying to amend the last commit by doing:
git --soft reset && git commit
However, the user can abort commit for some reason by providing an empty
log message, and that would leave the last commit undone, while the user
being completely unaware about what happened. Now if the user tries to
continue, by issuing "git rebase --continue" that squashes two previous
commits.
Signed-off-by: Dmitry Potapov <dpotapov@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
